Professional Documents
Culture Documents
Image Mosaicing: Presented by
Image Mosaicing: Presented by
Mosaicing
Its techniques, algorithms, implementations and
applications
Presented By:
Sk Saddam Ahmed
Roll No: 02
M.Tech in Computer Science & Engg
1
Contents
Slide Number
2
What is Mosaic and Mosaicing?
• “Mosaic“ originates from an old Italian word “mosaico” which
means a picture or pattern produced by arranging together small
pieces of stone, tile, glass, etc.
Example:
3
Image Mosaicing
• Many a time, it may not be possible to capture the complete image of a
large document in a single exposure as most image-capturing media work
with documents of definite size.
• In such cases, the document has to be scanned part by part producing split
images. Thus, document image analysis and processing require Mosaicing
of the split images to obtain a complete final image of the document.
4
Why We Need Image
Mosaicing?
• There are situations where it is not possible to capture large
documents with the given imaging media such as scanners or
copying machines in a single stretch because of their inherent
limitations.
• Image mosaicing not only allow you to create a large field of view
using normal camera, the result image can also be used for texture
mapping of a 3D environment such that users can view the
surrounding scene with real images.
5
Image Mosaicing Model
Input Images
Feature Extraction
Image Registration
Homographic
Refinement
Output Mosaic
image Output
Mosaic image
6
Image Mosaicing Model(Contd..)
1) Feature Extraction
The first step in image mosaic process is feature detection. Features are the elements in the
two input images to be matched. For images to be matched they are taken inside an image
patches. These image patches are groups of pixel in images. Patch matching is done for the
input images .
2) Image Registration
Image registration is the process of aligning two or more images of the same scene taken at
different times. It geometrically aligns two images—the reference and sensed images. This
process is needed in various computer vision applications like motion analysis, change detection,
image fusion etc.
Reference Image Sensed Image Registered Image
7
Image Mosaicing Model(Contd..)
3) Homographic Refinement
Homography is mapping between two spaces which often used to represent the correspondence
between two images of the same scene. It’s widely useful for images where multiple images are
taken from a rotating camera having a fixed camera centre ultimately warped together to produce
a panoramic view.
8
Image Mosaicing Model(Contd..)
5) Image Blending
The final step is to blend the pixels colours in the overlapped region to avoid the seams. Simplest
available form is to use feathering ,which uses weighted averaging colour values to blend the
overlapping pixels.
9
Algorithms for Image
Mosaicing
10
Unidirectional Algorithm
Here we present an algorithm to obtain a mosaiced image from its split images that has
the time complexity O(n2) based on comparing the values of pixels in each column of the
split images of a large document.
Algorithm:
11
Drawback of Unidirectional Algorithm
The experimental results of this algorithm do not conform to the expected results. The input
split image 1 (figure a) contains more overlapping region compared to split image 2 (figure
b). In such situations this algorithm fails to give the complete overlapping region present in
both the images, which is the major drawback of this method.
12
Bidirectional Algorithm
This method overcomes the drawback of the first algorithm, but has the same time complexity. This
algorithm is thus just an extension of the previous algorithm. The method scans the split images
from right to left as well as left to right, whereas in previous Algorithm scanning of the image takes
place only from left to right to identify the overlapping region in the split images.
Algorithm:
13
Mosaicing of two split images with O(n2) time complexity by back
tracking.
Results
Image 1 Image 2
Mosaic
Image
Image 3
Image 1 Image 2
Mosaic
Image
Image 3
14
Limitations
• The methods work fine for all types of documents but they
consume time.
15
Application
• Constructing high resolution images that cover an unlimited field
of view using inexpensive equipment.
16
References
• [Hartley] Hartley, R. & Zisserman, A. (2000) Multiple View Geometry{Cambridge University
Press, UK.
• [Shum] Shum, H. & Szeliski, R. (1998) Construction and refinement of panoramic mosaics with
global and local alignment. IEEE Int'l Conf. Computer Vision, pp. 953-958.
• [Faugeras] Zoghlami, I. & Faugeras,O. & Deriche,R. (1997) Using geometric corners to build a
2d mosaic from as set of images.Computer Vision and Pattern Recognition, pp 421-425.
• [Zhang] Zhang, Z.& Deriche, R. & Faugeras, O & Luong, Q. A robust technique for matching two
uncalibrated images through the recovery of the unknown epipolar geometry (1995) Artificial
Intelligence Journal, 78:87-119, October 1995
• [Harris] Harris, C. & Stephens, M. A combined corner and edge detector.(1998) Proc. of 4th
Alvey Vision Conf.,147-151.
• [Szeliski] Szeliski, R. Image Mosaicing for Tele-Reality Applications.(1994). Digital Equipment
Corporation, Cambridge, USA.
• [Davis] Davis, J. Mosaics of scenes with moving objects.(1998).Computer Vision and Pattern
Recognition
• [Capel] Capel,D & Zisserman,A. Automated mosaicing with superresolutionzoom.
(1998).Computer Vision and Pattern Recognition.
17
Thank You
18